Load data¶
In [1]:
# Load data
import glob
import cv2
vehicles = glob.glob('data/vehicles/**/*.png')
non_vehicles = glob.glob('data/non-vehicles/**/*.png')
cars = list(map(lambda x: cv2.cvtColor(cv2.imread(x), cv2.COLOR_BGR2RGB), vehicles))
non_cars = list(map(lambda x: cv2.cvtColor(cv2.imread(x), cv2.COLOR_BGR2RGB), non_vehicles))
print(len(cars))
print(len(non_cars))
Examine data images¶
In [2]:
# Explore data
import numpy as np
import matplotlib.pyplot as plt
%matplotlib inline
for i in range(20):
car = cv2.imread(vehicles[np.random.randint(0,len(vehicles))])
non_car = cv2.imread(non_vehicles[np.random.randint(0,len(non_vehicles))])
# Print example of car and non-car
if i == 1:
cv2.imwrite('output_images/Car_Image.jpg', car)
cv2.imwrite('output_images/Non-Car_Image.jpg', non_car)
figure, axes = plt.subplots(1,2, figsize=(10, 10))
figure.tight_layout()
#axes[0].set_axis_off()
axes[0].imshow(cv2.cvtColor(car, cv2.COLOR_BGR2RGB))
axes[0].set_title('Car Image', fontsize=30)
axes[1].imshow(cv2.cvtColor(non_car, cv2.COLOR_BGR2RGB))
axes[1].set_title('Non-Car Image', fontsize=30)
plt.subplots_adjust(left=0., right=1, top=0.9, bottom=0.)
Feature extraction functions from class¶
Spatial binning of color¶
In [3]:
def bin_spatial(img, size=(32, 32)):
# Use cv2.resize().ravel() to create the feature vector
features = cv2.resize(img, size).ravel()
# Return the feature vector
return features
Histogram of colors¶
In [4]:
def color_hist(img, nbins=32, bins_range=(0, 256)):
# Compute the histogram of the color channels separately
channel1_hist = np.histogram(img[:,:,0], bins=nbins, range=bins_range)
channel2_hist = np.histogram(img[:,:,1], bins=nbins, range=bins_range)
channel3_hist = np.histogram(img[:,:,2], bins=nbins, range=bins_range)
# Concatenate the histograms into a single feature vector
hist_features = np.concatenate((channel1_hist[0], channel2_hist[0], channel3_hist[0]))
# Return the individual histograms, bin_centers and feature vector
return hist_features
HOG features¶
In [5]:
from skimage.feature import hog
# Define a function to return HOG features and visualization
def get_hog_features(img, orient, pix_per_cell, cell_per_block,
vis=False, feature_vec=True):
# Call with two outputs if vis==True
if vis == True:
features, hog_image = hog(img, orientations=orient, pixels_per_cell=(pix_per_cell, pix_per_cell),
cells_per_block=(cell_per_block, cell_per_block), block_norm= 'L2-Hys',
transform_sqrt=True,
visualise=vis, feature_vector=feature_vec)
return features, hog_image
# Otherwise call with one output
else:
features = hog(img, orientations=orient, pixels_per_cell=(pix_per_cell, pix_per_cell),
cells_per_block=(cell_per_block, cell_per_block), block_norm= 'L2-Hys',
transform_sqrt=True,
visualise=vis, feature_vector=feature_vec)
return features
In [37]:
# Try cv2.HOGDescriptor
def get_cv2_hog_features(image, winSize = (64,64),\
blockSize = (16,16),\
blockStride = (8,8),\
cellSize = (8,8),\
nbins = 9, \
cspace = 'YCrCb'):
if cspace != 'RGB':
if cspace == 'HSV':
feature_image = cv2.cvtColor(image, cv2.COLOR_RGB2HSV)
elif cspace == 'LUV':
feature_image = cv2.cvtColor(image, cv2.COLOR_RGB2LUV)
elif cspace == 'HLS':
feature_image = cv2.cvtColor(image, cv2.COLOR_RGB2HLS)
elif cspace == 'YUV':
feature_image = cv2.cvtColor(image, cv2.COLOR_RGB2YUV)
elif cspace == 'YCrCb':
feature_image = cv2.cvtColor(image, cv2.COLOR_RGB2YCrCb)
else: feature_image = np.copy(image)
hog = cv2.HOGDescriptor(winSize,blockSize,blockStride,cellSize,nbins)
return np.ravel(hog.compute(feature_image))
Function to extract and combine all three features above¶
In [6]:
def extract_features(image):
# Extract parameters.
cspace = 'YCrCb' # Can be RGB, HSV, LUV, HLS, YUV, YCrCb
orient = 8
pix_per_cell = 8
cell_per_block = 2
hog_channel = 'ALL' # Can be 0, 1, 2, or "ALL"
spatial_size = (16, 16)
hist_bins = 32
hist_range = (0, 256)
if cspace != 'RGB':
if cspace == 'HSV':
feature_image = cv2.cvtColor(image, cv2.COLOR_RGB2HSV)
elif cspace == 'LUV':
feature_image = cv2.cvtColor(image, cv2.COLOR_RGB2LUV)
elif cspace == 'HLS':
feature_image = cv2.cvtColor(image, cv2.COLOR_RGB2HLS)
elif cspace == 'YUV':
feature_image = cv2.cvtColor(image, cv2.COLOR_RGB2YUV)
elif cspace == 'YCrCb':
feature_image = cv2.cvtColor(image, cv2.COLOR_RGB2YCrCb)
else: feature_image = np.copy(image)
# Call get_hog_features() with vis=False, feature_vec=True
if hog_channel == 'ALL':
hog_features = []
for channel in range(feature_image.shape[2]):
hog_features.append(get_hog_features(feature_image[:,:,channel],
orient, pix_per_cell, cell_per_block,
vis=False, feature_vec=True))
hog_features = np.ravel(hog_features)
else:
hog_features = get_hog_features(feature_image[:,:,hog_channel], orient,
pix_per_cell, cell_per_block, vis=False, feature_vec=True)
# Apply bin_spatial() to get spatial color features
spatial_features = bin_spatial(feature_image, size=spatial_size)
# Apply color_hist() also with a color space option now
hist_features = color_hist(feature_image, nbins=hist_bins, bins_range=hist_range)
# Return list of feature vectors
return np.concatenate((spatial_features, hist_features, hog_features))
In [20]:
def extract_cv2hog_features(image):
cspace = 'YCrCb'
winSize = (64,64)
blockSize = (16,16)
blockStride = (8,8)
cellSize = (8,8)
nbins = 9
if cspace != 'RGB':
if cspace == 'HSV':
feature_image = cv2.cvtColor(image, cv2.COLOR_RGB2HSV)
elif cspace == 'LUV':
feature_image = cv2.cvtColor(image, cv2.COLOR_RGB2LUV)
elif cspace == 'HLS':
feature_image = cv2.cvtColor(image, cv2.COLOR_RGB2HLS)
elif cspace == 'YUV':
feature_image = cv2.cvtColor(image, cv2.COLOR_RGB2YUV)
elif cspace == 'YCrCb':
feature_image = cv2.cvtColor(image, cv2.COLOR_RGB2YCrCb)
else: feature_image = np.copy(image)
hog_features = []
hog_features.append(get_cv2_hog_features(feature_image, winSize,\
blockSize,\
blockStride,\
cellSize,\
nbins))
hog_features = np.ravel(hog_features)
# Apply bin_spatial() to get spatial color features
spatial_features = bin_spatial(feature_image, size=spatial_size)
# Apply color_hist() also with a color space option now
hist_features = color_hist(feature_image, nbins=hist_bins, bins_range=hist_range)
return np.concatenate((spatial_features, hist_features, hog_features))
Extract features¶
In [25]:
from sklearn.preprocessing import StandardScaler
from sklearn.model_selection import train_test_split
import time
cspace = 'YCrCb' # Can be RGB, HSV, LUV, HLS, YUV, YCrCb
orient = 8
pix_per_cell = 8
cell_per_block = 2
hog_channel = 'ALL' # Can be 0, 1, 2, or "ALL"
spatial_size = (16, 16)
hist_bins = 32
hist_range = (0, 256)
t=time.time()
#car_features = list(map(lambda car: extract_features(car), cars ))
car_features = list(map(lambda car: extract_cv2hog_features(car), cars ))
#notcar_features = list(map(lambda notcar: extract_features(notcar), non_cars ))
notcar_features = list(map(lambda notcar: extract_cv2hog_features(notcar), non_cars ))
t2 = time.time()
print(round(t2-t, 2), 'Seconds to extract HOG features...')
# Create an array stack of feature vectors
X = np.vstack((car_features, notcar_features)).astype(np.float64)
# Define the labels vector
y = np.hstack((np.ones(len(car_features)), np.zeros(len(notcar_features))))
# Split up data into randomized training and test sets
rand_state = np.random.randint(0, 100)
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=rand_state)
# Fit a per-column scaler
X_scaler = StandardScaler().fit(X_train)
# Apply the scaler to X
X_train = X_scaler.transform(X_train)
X_test = X_scaler.transform(X_test)
print('Using:',orient,'orientations',pix_per_cell,
'pixels per cell and', cell_per_block,'cells per block')
print('Feature vector length:', len(X_train[0]))
Train SVM classifier¶
In [26]:
# Train SVM classifier
from sklearn.svm import LinearSVC
# Use a linear SVC
svc = LinearSVC()
# Check the training time for the SVC
t=time.time()
svc.fit(X_train, y_train)
t2 = time.time()
print(round(t2-t, 2), 'Seconds to train SVC...')
# Check the score of the SVC
print('Test Accuracy of SVC = ', round(svc.score(X_test, y_test), 4))
# Check the prediction time for a single sample
t=time.time()
n_predict = 10
print('My SVC predicts: ', svc.predict(X_test[0:n_predict]))
print('For these',n_predict, 'labels: ', y_test[0:n_predict])
t2 = time.time()
print(round(t2-t, 5), 'Seconds to predict', n_predict,'labels with SVC')
In [55]:
# Following suggestion and try another classifier to
# achieve at leat 99% accuracy
from sklearn.neural_network import MLPClassifier
# Try MLP classifier
mlp=MLPClassifier(random_state=777)
# Check the training time for the MLP
t=time.time()
mlp.fit(X_train, y_train)
t2 = time.time()
print(round(t2-t, 2), 'Seconds to train MLP...')
# Check the score of the SVC
print('Test Accuracy of MLP = ', round(mlp.score(X_test, y_test), 4))
# Check the prediction time for a single sample
t=time.time()
n_predict = 10
print('My MLP predicts: ', mlp.predict(X_test[0:n_predict]))
print('For these',n_predict, 'labels: ', y_test[0:n_predict])
t2 = time.time()
print(round(t2-t, 5), 'Seconds to predict', n_predict,'labels with MLP')
Visualize Histogram Oriented Gradients (HOG)¶
Cars¶
In [27]:
for i in range(20):
car = cv2.imread(vehicles[np.random.randint(0,len(vehicles))])
_, car_hog = get_hog_features(car[:,:,2], orient, pix_per_cell, cell_per_block,
vis=True, feature_vec=True)
# Print HOG examples of car and non-car
if i == 1:
cv2.imwrite('output_images/Before_HOG.jpg', car)
cv2.imwrite('output_images/After_HOG.png', car_hog)
figure, axes = plt.subplots(1,2, figsize=(10, 10))
figure.tight_layout()
#axes[0].set_axis_off()
axes[0].imshow(car)
axes[0].set_title('Car Image', fontsize=30)
axes[1].imshow(car_hog)
axes[1].set_title('HOG Image', fontsize=30)
plt.subplots_adjust(left=0., right=1, top=0.9, bottom=0.)
Non-cars¶
In [28]:
for i in range(20):
non_car = cv2.imread(non_vehicles[np.random.randint(0,len(non_vehicles))])
_, noncar_hog = get_hog_features(non_car[:,:,2], orient, pix_per_cell, cell_per_block,
vis=True, feature_vec=True)
figure, axes = plt.subplots(1,2, figsize=(10, 10))
figure.tight_layout()
#axes[0].set_axis_off()
axes[0].imshow(non_car)
axes[0].set_title('Non-car Image', fontsize=30)
axes[1].imshow(noncar_hog)
axes[1].set_title('HOG Image', fontsize=30)
plt.subplots_adjust(left=0., right=1, top=0.9, bottom=0.)
Sliding window and search for car¶
In [29]:
import matplotlib.image as mpimg
# Here is your draw_boxes function from the previous exercise
def draw_boxes(img, bboxes, color=(0, 0, 255), thick=6):
# Make a copy of the image
imcopy = np.copy(img)
# Iterate through the bounding boxes
for bbox in bboxes:
# Draw a rectangle given bbox coordinates
cv2.rectangle(imcopy, bbox[0], bbox[1], color, thick)
# Return the image copy with boxes drawn
return imcopy
# Define a function that takes an image,
# start and stop positions in both x and y,
# window size (x and y dimensions),
# and overlap fraction (for both x and y)
def slide_window(img, x_start_stop=[None, None], y_start_stop=[None, None],
xy_window=(64, 64), xy_overlap=(0.5, 0.5)):
# If x and/or y start/stop positions not defined, set to image size
if x_start_stop[0] == None:
x_start_stop[0] = 0
if x_start_stop[1] == None:
x_start_stop[1] = img.shape[1]
if y_start_stop[0] == None:
y_start_stop[0] = 0
if y_start_stop[1] == None:
y_start_stop[1] = img.shape[0]
# Compute the span of the region to be searched
xspan = x_start_stop[1] - x_start_stop[0]
yspan = y_start_stop[1] - y_start_stop[0]
# Compute the number of pixels per step in x/y
nx_pix_per_step = np.int(xy_window[0]*(1 - xy_overlap[0]))
ny_pix_per_step = np.int(xy_window[1]*(1 - xy_overlap[1]))
# Compute the number of windows in x/y
nx_buffer = np.int(xy_window[0]*(xy_overlap[0]))
ny_buffer = np.int(xy_window[1]*(xy_overlap[1]))
nx_windows = np.int((xspan-nx_buffer)/nx_pix_per_step)
ny_windows = np.int((yspan-ny_buffer)/ny_pix_per_step)
# Initialize a list to append window positions to
window_list = []
# Loop through finding x and y window positions
# Note: you could vectorize this step, but in practice
# you'll be considering windows one by one with your
# classifier, so looping makes sense
for ys in range(ny_windows):
for xs in range(nx_windows):
# Calculate window position
startx = xs*nx_pix_per_step + x_start_stop[0]
endx = startx + xy_window[0]
starty = ys*ny_pix_per_step + y_start_stop[0]
endy = starty + xy_window[1]
# Append window position to list
window_list.append(((startx, starty), (endx, endy)))
# Return the list of windows
return window_list
In [64]:
#def search_windows(img, windows, clf, scaler, color_space,
# spatial_size, hist_bins,
# hist_range, orient,
# pix_per_cell, cell_per_block,
# hog_channel, spatial_feat,
# hist_feat, hog_feat):
def search_windows(img, windows, clf, scaler):
#1) Create an empty list to receive positive detection windows
on_windows = []
#2) Iterate over all windows in the list
for window in windows:
#3) Extract the test window from original image
test_window = cv2.resize(img[window[0][1]:window[1][1], window[0][0]:window[1][0]], (64, 64))
#4) Extract features for that window using single_img_features()
#features = extract_features(test_window)
features = extract_cv2hog_features(test_window)
#5) Scale extracted features to be fed to classifier
img_features = scaler.transform(np.array(features).reshape(1, -1))
#6) Predict using your classifier and distance away from decision boundary
#prediction = svc.predict(img_features)
#confidence = svc.decision_function(img_features)
#
# If positive (prediction == 1) and above confidence threshold
# then save the window
#confidence_threshold = 0.5
#if prediction == 1 and abs(confidence) > confidence_threshold:
# Add MLP classifier
if clf.predict_proba(img_features)[0][1] > 0.99:
on_windows.append(window)
#8) Return windows for positive detections
return on_windows
Test pipeline on test images¶
In [67]:
# Test pipeline on all test images
xy_window=(64, 64)
xy_overlap=(0.7, 0.7)
y_start_stop = [360, 700] # Min and max in y to search in slide_window()
test_images = glob.glob('./test_images/test*.jpg')
for fname in test_images:
test_img = cv2.cvtColor(cv2.imread(fname), cv2.COLOR_BGR2RGB)
draw_image = np.copy(test_img)
windows = slide_window(test_img, x_start_stop=[None, None], y_start_stop=y_start_stop,
xy_window=xy_window, xy_overlap=xy_overlap)
bounding_boxes = []
#bounding_boxes += (search_windows(test_img, windows, svc, X_scaler))
bounding_boxes += (search_windows(test_img, windows, mlp, X_scaler))
window_img = draw_boxes(draw_image, bounding_boxes, color=(0, 0, 255), thick=6)
# Print examples of sliding windows
if fname == './test_images/test1.jpg':
cv2.imwrite('output_images/Before_slidingWindows.jpg', test_img)
cv2.imwrite('output_images/After_slidingWindows.jpg', window_img)
#plt.imshow(window_img)
figure, axes = plt.subplots(1,2, figsize=(24, 9))
figure.tight_layout()
#axes[0].set_axis_off()
axes[0].imshow(test_img)
axes[0].set_title('Original Image', fontsize=30)
axes[1].imshow(window_img)
axes[1].set_title('With Windows', fontsize=30)
plt.subplots_adjust(left=0., right=1, top=0.9, bottom=0.)
Multiple Detections & False Positives¶
In [34]:
import matplotlib.image as mpimg
import matplotlib.pyplot as plt
import numpy as np
import pickle
import cv2
from scipy.ndimage.measurements import label
def add_heat(heatmap, bbox_list):
# Iterate through list of bboxes
for box in bbox_list:
# Add += 1 for all pixels inside each bbox
# Assuming each "box" takes the form ((x1, y1), (x2, y2))
heatmap[box[0][1]:box[1][1], box[0][0]:box[1][0]] += 1
# Return updated heatmap
return heatmap# Iterate through list of bboxes
def apply_threshold(heatmap, threshold):
# Zero out pixels below the threshold
heatmap[heatmap <= threshold] = 0
# Return thresholded map
return heatmap
def draw_labeled_bboxes(img, labels):
# Iterate through all detected cars
for car_number in range(1, labels[1]+1):
# Find pixels with each car_number label value
nonzero = (labels[0] == car_number).nonzero()
# Identify x and y values of those pixels
nonzeroy = np.array(nonzero[0])
nonzerox = np.array(nonzero[1])
# Define a bounding box based on min/max x and y
bbox = ((np.min(nonzerox), np.min(nonzeroy)), (np.max(nonzerox), np.max(nonzeroy)))
# Draw the box on the image
cv2.rectangle(img, bbox[0], bbox[1], (0,0,255), 6)
# Return the image
return img
In [76]:
def find_cars(img, xy_window, xy_overlap, y_start_stop, scale, clf, X_scaler):
# Extract parameters.
cspace = 'YCrCb' # Can be RGB, HSV, LUV, HLS, YUV, YCrCb
scale = scale
xy_window = xy_window
xy_overlap = xy_overlap
y_start_stop = y_start_stop
bboxes = []
windows = slide_window(img, x_start_stop=[None, None], y_start_stop=y_start_stop,
xy_window=xy_window, xy_overlap=xy_overlap)
bboxes += (search_windows(img, windows, clf, X_scaler))
return bboxes
In [123]:
# Apply multiple sliding windows
test_img = cv2.cvtColor(cv2.imread('./test_images/test1.jpg'), cv2.COLOR_BGR2RGB)
draw_image = np.copy(test_img)
boundingboxes = []
xy_window=(64, 64)
xy_overlap=(0.8, 0.8)
y_start_stop = [350, 550]
scale = 2.0
boundingboxes.append(find_cars(test_img, xy_window, xy_overlap, y_start_stop, scale, mlp, X_scaler))
xy_window=(64, 64)
xy_overlap=(0.85, 0.85)
y_start_stop = [400, 500]
scale = 2.0
boundingboxes.append(find_cars(test_img, xy_window, xy_overlap, y_start_stop, scale, mlp, X_scaler))
xy_window=(128, 128)
xy_overlap=(0.85, 0.85)
y_start_stop = [450, 550]
scale = 2.0
boundingboxes.append(find_cars(test_img, xy_window, xy_overlap, y_start_stop, scale, mlp, X_scaler))
xy_window=(64, 64)
xy_overlap=(0.85, 0.85)
y_start_stop = [450, 600]
scale = 2.0
boundingboxes.append(find_cars(test_img, xy_window, xy_overlap, y_start_stop, scale, mlp, X_scaler))
xy_window=(128, 128)
xy_overlap=(0.85, 0.85)
y_start_stop = [350, 650]
scale = 2.0
boundingboxes.append(find_cars(test_img, xy_window, xy_overlap, y_start_stop, scale, mlp, X_scaler))
bboxes = [box for boxlist in boundingboxes for box in boxlist]
window_img = draw_boxes(draw_image, bboxes, color=(0, 0, 255), thick=6)
plt.figure(figsize=(10,10))
plt.imshow(window_img)
Out[123]:
In [124]:
# Apply heatmap
heatmap = np.zeros_like(test_img[:,:,0])
heatmap = add_heat(heatmap, bboxes)
plt.figure(figsize=(10,10))
plt.imshow(heatmap, cmap='hot')
Out[124]:
In [125]:
# Apply heatmap threshold
heatmap = apply_threshold(heatmap,2)
heatmap = np.clip(heat, 0, 255)
plt.figure(figsize=(10,10))
plt.imshow(heatmap, cmap='hot')
Out[125]:
In [126]:
# Find final boxes from heatmap using label function
labels = label(heatmap)
out_img = draw_labeled_bboxes(draw_image, labels)
plt.figure(figsize=(10,10))
plt.imshow(out_img)
Out[126]:
In [129]:
# Put all together and test on all test images
xy_window=(64, 64)
xy_overlap=(0.85, 0.85)
y_start_stop = [360, 700] # Min and max in y to search in slide_window()
test_images = glob.glob('./test_images/test*.jpg')
for fname in test_images:
test_img = cv2.cvtColor(cv2.imread(fname), cv2.COLOR_BGR2RGB)
draw_image = np.copy(test_img)
bounding_boxes = []
boundingboxes.append(find_cars(test_img, (64, 64), (0.8, 0.8), [350, 550], 2.0, mlp, X_scaler))
boundingboxes.append(find_cars(test_img, (64, 64), (0.85, 0.85), [400, 500], 2.0, mlp, X_scaler))
boundingboxes.append(find_cars(test_img, (128, 128), (0.85, 0.85), [450, 550], 2.0, mlp, X_scaler))
boundingboxes.append(find_cars(test_img, (128, 128), (0.85, 0.85), [350, 650], 2.0, mlp, X_scaler))
bboxes = [box for boxlist in boundingboxes for box in boxlist]
# Add heat to each box in box list
#heat = np.zeros(draw_image.shape)
heat = np.zeros_like(test_img[:,:,0])
heat = add_heat(heat,bboxes)
# Apply threshold to help remove false positives
heat = apply_threshold(heat,4)
# Visualize the heatmap when displaying
heatmap = np.clip(heat, 0, 255)
# Find final boxes from heatmap using label function
labels = label(heatmap)
out_img = draw_labeled_bboxes(draw_image, labels)
#window_img = draw_boxes(draw_image, hot_windows, color=(0, 0, 255), thick=6)
# Print examples of heatmaps and labels
if fname == './test_images/test1.jpg':
cv2.imwrite('output_images/Before_HeatmapAndLabels.jpg', test_img)
cv2.imwrite('output_images/After_HeatmapAndLabels.jpg', out_img)
figure, axes = plt.subplots(1,2, figsize=(24, 9))
figure.tight_layout()
#axes[0].set_axis_off()
axes[0].imshow(test_img)
axes[0].set_title('Original Image', fontsize=30)
axes[1].imshow(out_img)
axes[1].set_title('With Windows', fontsize=30)
plt.subplots_adjust(left=0., right=1, top=0.9, bottom=0.)
In [77]:
# Test find cars on all test images
#scale = 1.5
#xy_window=(64, 64)
#xy_overlap=(0.85, 0.85)
#y_start_stop = [360, 700] # Min and max in y to search in slide_window()
test_images = glob.glob('./test_images/test*.jpg')
for fname in test_images:
test_img = cv2.cvtColor(cv2.imread(fname), cv2.COLOR_BGR2RGB)
draw_image = np.copy(test_img)
bboxes = find_cars(test_img, (64, 64),(0.85, 0.85),\
[360, 700], 1.5, mlp, X_scaler)
# Add heat to each box in box list
heat = np.zeros(draw_image.shape)
heat = add_heat(heat,bboxes)
# Apply threshold to help remove false positives
heat = apply_threshold(heat,4)
# Visualize the heatmap when displaying
heatmap = np.clip(heat, 0, 255)
# Find final boxes from heatmap using label function
labels = label(heatmap)
out_img = draw_labeled_bboxes(draw_image, labels)
# Print examples of HOG sub-sampling
if fname == './test_images/test1.jpg':
cv2.imwrite('output_images/Before_find_cars.jpg', test_img)
cv2.imwrite('output_images/After_find_cars.jpg', out_img)
figure, axes = plt.subplots(1,2, figsize=(24, 9))
figure.tight_layout()
#axes[0].set_axis_off()
axes[0].imshow(test_img)
axes[0].set_title('Original Image', fontsize=30)
axes[1].imshow(out_img)
axes[1].set_title('With Windows', fontsize=30)
plt.subplots_adjust(left=0., right=1, top=0.9, bottom=0.)
HOG sub-sampling window search¶
In [46]:
# Define a single function that can extract features using hog sub-sampling and make predictions
def find_cars_with_hog_subsampling(img, svc, X_scaler):
# Extract parameters.
cspace = 'YCrCb' # Can be RGB, HSV, LUV, HLS, YUV, YCrCb
orient = 8
pix_per_cell = 8
cell_per_block = 2
cells_per_step = 1
hog_channel = 'ALL' # Can be 0, 1, 2, or "ALL"
spatial_size = (16, 16)
hist_bins = 32
hist_range = (0, 256)
ystart = 360
ystop = 700
scale = 1.5
bboxes = []
#draw_img = np.copy(img)
#img = img.astype(np.float32)/255
# apply color conversion if other than 'RGB'
if cspace != 'RGB':
if cspace == 'HSV':
feature_image = cv2.cvtColor(img, cv2.COLOR_RGB2HSV)
elif cspace == 'LUV':
feature_image = cv2.cvtColor(img, cv2.COLOR_RGB2LUV)
elif cspace == 'HLS':
feature_image = cv2.cvtColor(img, cv2.COLOR_RGB2HLS)
elif cspace == 'YUV':
feature_image = cv2.cvtColor(img, cv2.COLOR_RGB2YUV)
elif cspace == 'YCrCb':
feature_image = cv2.cvtColor(img, cv2.COLOR_RGB2YCrCb)
else: feature_image = np.copy(img)
ctrans_tosearch = feature_image[ystart:ystop,:,:]
if scale != 1:
imshape = ctrans_tosearch.shape
ctrans_tosearch = cv2.resize(ctrans_tosearch, (np.int(imshape[1]/scale), np.int(imshape[0]/scale)))
ch1 = ctrans_tosearch[:,:,0]
ch2 = ctrans_tosearch[:,:,1]
ch3 = ctrans_tosearch[:,:,2]
# Define blocks and steps as above
nxblocks = (ch1.shape[1] // pix_per_cell) - cell_per_block + 1
nyblocks = (ch1.shape[0] // pix_per_cell) - cell_per_block + 1
nfeat_per_block = orient*cell_per_block**2
# 64 was the orginal sampling rate, with 8 cells and 8 pix per cell
window = 64
nblocks_per_window = (window // pix_per_cell) - cell_per_block + 1
#cells_per_step = 2 # Instead of overlap, define how many cells to step
nxsteps = (nxblocks - nblocks_per_window) // cells_per_step
nysteps = (nyblocks - nblocks_per_window) // cells_per_step
# Compute individual channel HOG features for the entire image
hog1 = get_hog_features(ch1, orient, pix_per_cell, cell_per_block,feature_vec=False)
hog2 = get_hog_features(ch2, orient, pix_per_cell, cell_per_block,feature_vec=False)
hog3 = get_hog_features(ch3, orient, pix_per_cell, cell_per_block,feature_vec=False)
#hog1 = get_cv2_hog_features(ch1)
#hog2 = get_cv2_hog_features(ch2)
#hog3 = get_cv2_hog_features(ch3)
for xb in range(nxsteps):
for yb in range(nysteps):
ypos = yb*cells_per_step
xpos = xb*cells_per_step
# Extract HOG for this patch
hog_feat1 = hog1[ypos:ypos+nblocks_per_window, xpos:xpos+nblocks_per_window].ravel()
hog_feat2 = hog2[ypos:ypos+nblocks_per_window, xpos:xpos+nblocks_per_window].ravel()
hog_feat3 = hog3[ypos:ypos+nblocks_per_window, xpos:xpos+nblocks_per_window].ravel()
hog_features = np.hstack((hog_feat1, hog_feat2, hog_feat3))
#hog_features = get_cv2_hog_features(feature_image)
#hog_features = np.ravel(hog_features)
xleft = xpos*pix_per_cell
ytop = ypos*pix_per_cell
# Extract the image patch
subimg = cv2.resize(ctrans_tosearch[ytop:ytop+window, xleft:xleft+window], (64,64))
# Get color features
spatial_features = bin_spatial(subimg, size=spatial_size)
hist_features = color_hist(subimg, nbins=hist_bins)
# Scale features and make a prediction
test_features = X_scaler.transform(np.hstack((spatial_features, hist_features, hog_features)).reshape(1, -1))
test_prediction = svc.predict(test_features)
if test_prediction == 1:
xbox_left = np.int(xleft*scale)
ytop_draw = np.int(ytop*scale)
win_draw = np.int(window*scale)
#cv2.rectangle(draw_img,(xbox_left, ytop_draw+ystart),(xbox_left+win_draw,ytop_draw+win_draw+ystart),(0,0,255),6)
bboxes.append(((xbox_left, ytop_draw+ystart),(xbox_left+win_draw,ytop_draw+win_draw+ystart)))
#return draw_img
return bboxes
Video Implementation¶
In [74]:
from moviepy.editor import VideoFileClip
from IPython.display import HTML
# Define a function to process video
def process_video(img):
rgb_img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
draw_img = np.copy(img)
bboxes =find_cars(rgb_img, mlp, X_scaler)
heat = np.zeros(rgb_img.shape)
heatmap = add_heat(heat,bboxes)
# Integrate heat map over several (say 3) frames of video
heatmap_arr = []
if len(heatmap_arr) >= 3:
heatmap_arr = heatmap_arr[1:] # Discarding previous frame's heatmap
heatmap_arr.append(heatmap)
heatmap_avg = sum(heatmap_arr)/3
heatmap_frame = apply_threshold(heatmap_avg, 6)
labels = label(heatmap_frame)
return draw_labeled_bboxes(draw_img, labels)
In [75]:
video_input = './short_project_video.mp4'
video_output = 'output_images/VehicleDetection_video_out3.mp4'
video_clip = VideoFileClip(video_input)
processed_video = video_clip.fl_image(process_video)
%time processed_video.write_videofile(video_output, audio=False)